Med thumb creepy twitter study

Can you picture the Instagram feed of someone with an unquenching need for validation? Seems easy enough. What about the Facebook profile of one who doesn’t get the difference between typing in a URL and searching for weather.com? I bet yes.

One more: the feed of an insomniac tweeter? Apparently they, too, can now be profiled.

The insomniac Twitter user may not be as well-defined a stereotype, but researchers at Boston Children’s Hospital and Big Pharma manufacturer Merck are trying to flesh it out. In fact, they’ve begun to build a “digital phenotype” of insomnia using Twitter data, Medical News Today reported.

Their findings, published in the Journal of Medical Internet Research, describe some of the earliest formal research on the relationship between sleep disorders and social media use. Here’s how they built a profile of a twitter user with disordered sleep habits:

They started with data from two groups of twitter users. The first group of users’ tweets contained sleep-related language (insomnia), hashtags (#teamnosleep) or names of sleep medication (Ambien). The other group didn’t tweet about sleep.

While the users remained anonymous, researchers collected information on their age, location, daily and longer-term twitter activity, followers and favorited tweets. Researchers also categorized each tweet by time of day and tone (negative, positive, neutral).

The profile:

  • Compared to a well-rested tweeter, the average sleep-starved user has fewer followers and follows fewer people.
  • While the insomniac user has been active for a longer period of time, they are overall less active on a daily basis except during sleeping hours.
  • They’re also more likely to post negative tweets, suggesting a possible increased risk for psychosocial issues.

At this point, these findings make for interesting dinner conversation — and little more. It’s a preliminary study that compels further research.

The greater question is: Have scientists and corporate behemoths used your clever hashtags to build a phenotype of disease? Maybe next they’ll diagnose your paranoia.